App Monetization

App Monetization, Research

Header image - the SOOMLA mobile monetization report for Q2 2017 is full of insights about ad revenue in mobile apps

We are super excited to announce our insights report today. We started this practice in 2015 with reports that were more focused on in-app purchase based monetization but this one is all about insights related to monetization through ad revenues. The report explores domains that have never been explored before so lots of interesting insights on this one.

You are welcome to download the report through this link or via the banner to the right.

Would also be great if you can help us spread the word by sharing my post on Linkedin.

Linkedin post about mobile monetization report - q2 2017

 

Feel free to share:
Analytics, App Monetization

A browser screen with an eye representing impression, 29 percent written next to 1st impression, also the word volumes is written and eCPM next to a bar chart

About a week ago a friend asked me for a piece of information that should probably interest many others as well. He wanted to know how for rewarded video ads – how many impressions are first impressions vs. second impressions vs. third and so forth. In other words, he wanted to know how big of a deal first impressions are.

Impressions can be analyzed according to their sequence

To understand his quesion, we first need to understand the basics of user interaction with ads. When it comes to linear ad formats such as interstitials, video and rewarded video a user can only watch one at a time. This means that ad impressions have sequence and can be put in order. The first impression a user watches in a given day is considered the most fresh advertising experience he will get and typically yields more for the publisher while providing more value for the advertiser. A user might watch more impressions, a 2nd impression, a 3rd impression and so on. Checking the distribution of ads according to their sequecne means checking how many impressions are first impressions vs. second impressions vs. 3rd and so on and what percentage of the total volume each sequence position gets.

Results – the first 2 impressions give 46% of the volume

The results we found are presented in the chart below. We aggreagated data across all the apps using SOOMLA TRACEBACK and combined the results to a single chart. We excluded apps with less than 100,000 monthly impressiosn. The chart below represents the average with equal weights. In other words, the patterns of apps with high volume and the patterns of apps with smaller volume are equally represented.

Bar chart representing the impression volume for every impression sequence place. The logo of SOOMLA TRACEBACK is also shown

The full data can also be viewed in this table. We also included the minimal and maximal numbers accross all apps.

Impression Min Avg Max
1 13.6% 29.1% 48.3%
2 13.3% 17.4% 22.5%
3 11.6% 12.6% 13.8%
4 6.9% 9.5% 11.4%
5 4.2% 7.9% 9.8%
6 2.5% 6.3% 9.3%
7 1.6% 5.3% 8.9%
8 1.0% 4.5% 7.9%
9 0.6% 4.0% 7.7%
10 0.4% 3.5% 7.2%

First impressions matter

We already talked about the importance of first impressions from an eCPM standpoint in this article and also in this one. According to the data presented here, firstl impressions in rewarded video also matter because they represents a big chunk of the volume.

Feel free to share:
App Monetization

Ecpm of new users could be 2x higher compared to loyal users in mobile apps and mobil games

We already wrote here a few times about eCPM decay and some of our tips were also quoted in other places. In this post we are going to talk about another type of eCPM decay – the one that is rarely mentioned. I’m talking about the trend of eCPM going down as a function of how long the user was retained in the game as opposed to the decay that happens as a result of high frequency of ads during the day.

First Test – How eCPM behaves over the life of a user

In this test we looked for users who started playing the game in a certain month and than checked their eCPM in that same month versus the eCPM in the following month and the third month. We did this test across many games to make sure the results are not isolated to a single game. In this chart below you can see the average values, the maximal ones (the game with the highest rate in that month) and the minimal ones (the game with the lowest rate in that month) across all the games we tested. Note that this test was done only with US based users and only in the following ad formats: Offer walls, Rewarded Videos and Interstitials.

Ecpm decay over time in different games showing the ecpm of users in theif first month, 2nd month and 3rd month since started playing the mobil game

There is clearly a trend here. The eCPM is going down the longer the user is retained in the game. In fact, new users can have 2x or more the eCPM of loyal users. We can attempt to explain this finding of course. One assumption is that the same behavior pattern that impacts eCPM decay also comes into play here. Users tend to grow tired of advertising. However, here the situation is a bit different. Consider the case of a user who downloaded a new game this month but might also downloaded another game 3 months ago. It’s the same user so why is he responding better to ads in the new game he downloaded vs. the older game? The answer could be that the user gets tired of ads in a given context seperately. He might learn where the ads are placed and his brain is getting trained better to ignore them. It will be interesting to see what happens if we mix up the ad placements for loyal users to see if we can engage them with the ads again.

Second Test – Does it matter where the user came from?

Here, we tried to see if a user that came organically behaves differently compared to a user that came through paid UA or cross promotion. We compared only for US based users – here is what we found.
Ecpm for users who came through different channels

So it looks like the Cross-promo traffic had very high eCPMs in the first month. Paid installs that came from Facebook also appeared higher than Organic. However, the drastic difference in the eCPM of the usres in the 1st month almost vanished when looking at the the 2nd and 3rd month. Specifically, the cross-promo installs were lower compared to organic installs in the 3rd month. In general, the eCPMs converge to the same levels almost. It seems that the impact of the source of the user only lasts for the first month and after that month the user ‘forgets’ where he came from and users behave in a similar fashion. It’s possible that users who came from an ad into your game are more likely to respond to ads in your game. The fact that the impact only lasts for 1 month could potentially be explained by users response to ads is a temporary behavior and not a long lasting behavior pattern.

Third Test – Do we see the same trend across all ad formats?

We wanted to see if all ad formats behave the same way when it comes to this type of eCPM decay. Do users lose their interest in rewarded videos the same way as they do with interstitials? We compared 3 ad formats and this time we compared not just US traffic but we allowed international traffic. To make it easier to follow we indexed the results so they all fit in the same scale.
Ecpm of users across different formats as a function of how much time they were retained in the game

It’s easy to notice that the findings are consistent across all ad formats we tested. We didn’t check banners and native ads in this study. It’s possible we will do another post specifically focused on that.

Optimizing for the long retained users

One conclusion from this data is that there should be opportunities to better serve ads for loyal users so they monetize better. Here are some ideas to consider specifically for this segment:

  • Serving ads through SSPs – these ads come with an upfront bid price and are less influenced by users’ ad engagement
  • Closing fixed CPM deals for this segment
  • Mixing it up – changing the placements for user who have been playing the game long enough

The impact on LTV calculations

These findings might also impact how companies think about LTV prediction. Many LTV models assume that eCPMs and ARPDAU are not influenced by the amount of time the user played the game. If your existing model is predicting LTV based on the 1st month’S eCPM the actual result might be worse than the predicted LTV.

What about Apps

While the reseacrch was focused on games only we expect that to find the same patterns in Apps. At least that is true for the formats we checked: Rewarded videos, Offer walls and Interstitials.

Feel free to share:
Analytics, App Monetization

Measuring and optimizing opt in rate in rewarded video

One of the areas in which a company can drastically increase ad revenues with a relatively low effort is through the optimization of opt-in ratios to rewarded videos. Here we will show how to track and optimize opt-in rates with SOOMLA TRACEBACK.

What is opt in rate and why does it matter

Rewarded video is a unique format in the sense that it’s not forced on users. The game is offering some reward in return for watching a video and the user can accept the offer or not. These offers can be made with a pop up message, a button on different screen in the app or sometimes by replacing a call to action that would normally be prompting the users to pay. Regardless of the offer method, the users can accept or decline. The number of users who decide to take the offer is often called engaged users and the dividing them by the total number of active users is considered the opt-in rate.
One thing that is clear is that users who don’t engage with the rewarded video don’t contribute any revenue so by increasing the opt-in rate we are making the pool of monetizing users bigger. It is also known that users monetize best in their first impression and so getting more users to opt-in means you are getting a lot more of those valuable 1st impressions. Our experience has shown that increasing the opt-in ratio by x% often translates to a similar increase in the total ad-revenue.

Measuring the opt in with SOOMLA TRACEBACK

One of the easiest ways to measure the opt in rate is to use the TRACEBACK platform. You can see your overall opt in rates and number of engaged users but you can also look at specific segments and breakdowns across these dimensions:

  • Countries
  • Platform/OS
  • Versions of your app
  • Traffic sources
  • Date ranges

Looking at specific segments allows you to find improvement opportunities. The way to spot these is simple – a low opt in rate means there is a room to grow it.

What is a good opt in rate

Depending on your game of course and how well you are doing with IAP monetization you can reach as high as 80% opt in rate according to this study by Unity Ads. However, apps that focus more on IAP would be smart to first convert the users into payers and only then try to push them harder to videos. From this reason we should look at the opt in rate on a cohorted basis and set different goals depending on the lifetime of the users. These are good benchmarks:

  • 1st month – 20%
  • 2nd month – 50%
  • 3rd month – 60%

Optimizing opt in rates

Once you have identified a segment that falls below the target opt in rate, you can use TRACEBACK to optimizie it. One way to do it is to track the opt in rate in different versions of your app. Whenever you launch a new version you can immediatly compare the opt in ratio to the one of the previous version and check if you are moving towards the goal or away from it.

Example – Optimizations results in higher opt in rate in later versions 

Here is an export from our dashboard into Excel showing the opt-in rates in different version of the app.

While comparing opt in rates between different versions is easy to do and comes as a built in feature of the platform, a better approach would be to use TRACEBACK alongside an a/b testing tool. This allows customers to compare between different versions and configurations simulatnously in a randomized testing environment. TRACEBACK will present the opt in rate for each testing group in the dashboard so you can easily compare and pick the winning configuration.

Feel free to share:
App Monetization

IMG_4102One of the things that were a part of mobile apps since the early days is the REMOVE ADS button. The idea is simple – ads generate low amounts of revenue per user and getting $0.99 or $1.99 from them is better from the app publisher stand point.

Not showing ads to payers is the standard practice

Even in games that don’t have a specific purchase option around removing ads it became a standard practice to not show ads to depositors. This is based on the same approach that ads yield low amounts of revenue while purchases yield higher amounts.

Rethink what you know – ad whales exist

In recent posts we covered the existence of ad whales. Individuals who generate large amounts of ad revenues for their app publishers. Here is a user who generated $74 in ad revenue in November, and this user generated $52 in December. While these levels of revenue per user are quite rare for ad monetization, they are also quite rare when it comes to in-app purchases.

How many users generate enough ad revenue to level with payers

If we consider how much revenue is generated by a payer – the minimum is $0.7. The lowest purchase by a user is $0.99 and given that Apple and Google take a cut of 30% the publisher gets 70 cents.
Based on the data SOOMLA Traceback is collecting we can check how many users go over the point. How many monetize with ads at least to the same level as payers. The result is that in some games that relay heavily on ads it’s more than 10% of the user base. This is higher than a normal conversion rate to payers. We can also check how many users went over $3.5 which is the publisher share of a more $4.99 purchase by a user. The result is that it’s over 2% in some games.

Rewarded videos offer incentives to users

Let’s start thinking about a different approach. Should we allow any type of advertising to people who paid? One area to consider is the type of advertising in question. Ads that may annoy a paying user could be a bad choice from a user experience perspective but what about incentivized formats such as offer walls and rewarded videos. These formats are loved by users so the question becomes more about optimizing the revenues.

Option 1 – reversed approach

Let’s imagine for a second a complete mirror image of the “no ads for payers” approach. What this means is that we set a threshold of $0.7 and the users who have made at least $0.7 in ad revenue are considered ad-whales. Once we classified someone as an ad-whale, we don’t allow him to make purchases in the game. That would be the reversed approach to the “no ads for payers” approach. If it sounds silly to you – it’s because it is silly. Blocking someone from paying in a game is just nonsense but so is the “no ads for payers” approach. Why block someone from making revenue for you through watching ads?

Option 2 – balanced approach

A more reasonable approach to the problem is to simply allow users constent access to all methods of getting benefits. A user can get benefits by buying them, by watching video ads, or by taking on offers. Since the payout of a video view by a user is normally determined in retrospect, the publisher could apply a model where the rewards are dynamic based on the past payouts received for that user. If such a model is implemented, the publisher can guarentee that the price of getting the benefit is balanced across the different methods the user has for getting them. For example, if the eCPM of a user starts falling after a while, his rewards for watching videos will decrease and he will be more inclined to make purchases. If however, the eCPMs for a specific users are growing over time, the rewards he will get from watching videos will increase and he will have more motivation to keep watching them as opposed to buying something.

Ad measurement tools are becoming a must have

This type of innovative monetization strategies are becoming critical for the survival of game studios. We covered before the increase in CPI rates and how companies needs to adapt to stay relevant. Advanced segmentation and monetization measurement tools that can find the ad whales segment for you are becoming a must have in today’s mobile eco-system.

 

If you want to start measuring your monetization and find ad whales you should check out SOOMLA Traceback – Ad LTV as a Service.

Learn More

Feel free to share:
App Monetization

Kate Uptown is starring the Machine Zone (MZ) ads for their Game of War which has been advertised heavily in the last 24 monthsDemand diversity is a topic not many people discuss in the mobile game monetization forums. To understand it let’s think about the journey or a user through our app. The first time the user watches an ad, the mediation will check what is the ad-network that is on the top of the waterfall today and will have that network serve the ad. The network will normally try to serve the highest yielding campaign they have – why don’t we call the app in the ad Mobile Assault – it will help us refer to it later. In many cases, this user will see the same ad over and over again in the same day and more time in the next day. Having a user see an ad for 100 times these days is not uncommon. This is the demand diversity problem I’m referring to.

Why demand diversity is important

From a user perspective, seeing the same ad over and over is a poor user experience. The first time you are seeing an ad, it could be interesting, cool or even funny. If you have seen the new Clash Royal ads, they are quite amusing. However, nothing is interesting, cool or funny after you have watched it a dozen times already. At that point, it is just annoying.
From an ad effectiveness perspective, showing the same ad over and over is a bad choice as well. It leads to banner blindness so users stop noticing the ad. Most ads today are shown with the purpose of creating installs for tha advertised app and blindness leads to low click rates and conversion rates so less installs are generated.

The business models determine who takes the risk

One of axioms of online advertising is the chart below. Basically it says:

  • In a CPM model the risk is on the advertiser side while the publisher has guarnteed income
  • CPC is the middle ground
  • In a CPA/CPI model the risk is on the publisher side while the advertiser has guarnteed outcome

Illustrative chart showing the risk levels for publishers and advertisers based on the selected business model: CPM, CPC or CPI

The mobile advertising industry today is mostly driven by the CPI model which is a form of CPA meaning that the publishers assume most of the risk. They place ads in their apps hoping to get paid but their monetization is driven by whether or not the users ended up taking additional actions outside of their apps.
So now that we established who has the risk, we also know how is the one that gets heart from the situation. Users who watch the same ad over and over again become blind to it and the publishers’ monetization levels are getting hurt.

Risk and data are normally aligned

In most business situations, the party who is willing to takes the risks is the one with better tools to assess it and mitigate it. For example, in a CPM model, the advertiser assume the risk but they demand transparency about where their ads are being placed and have tools to measure the performance. In mobile app monetization however, the publishers are the one assuming the risks but they are doing so with complete lack of data or measurement tools. More specifically, the publishers are the ones that get hurt from the lack of demand diversity but they actually have no way to measure and manage it.

Mediation platforms are also left in the dark

The parties that are in the perfect position to be the police of demand diversity are the mediation platforms. Publishers are trusting the mediation companies to act as their agents and help them manage things of this sort using their ad-tech expertiese. The problem is that mediation companies are also in the dark about what ad is being shown to the user. They simply call the ad-network SDK as a black box that shows ads but they don’t get any information out.

Ad networks only see their own ads

The only type of company that has information about what ads are shown to the user are the ad-networks. The problem, however, is that each ad-network is only aware of what ads they show. Instead of collaborating and sharing this data between them and be part of the solution they are part of the problem since an ad-network that is not aware of what other ad-networks are showing is likely to show the same popular advertiser again to the user.

Choosing ad networks smartly

App companies often tend to choose ad-networks based on rumors of their projected CPMs or based on how well it worked for their friends. Often, one ad-network will seem better than another in the eyes of the publishers due to their presence in shows and their general brand perception. However, choosing 4 networks that are practically representing the same demand menas making the problem worse. It’s common to see a rewarded video stack that includes Supersonic/Ironsource as the mediation in addition to Vungle, Adcolony, UnityAds and Chartboost as the ad-networks. These networks are considered the best when it comes to rewarded videos for mobile games. The problem here is that thery are all bringing similar types of ads. The chances of a user seeing the same ad over and over again is much higher like that. A smarter strategy for selecting ad partners is to try and figure out how to diversify. SSPs can often bring more diversification through access to exchanges and there are also companies like Mediabrix who focus only on bringing brand advertising.

Diversifying through blacklists

Most ad-networks supports blacklists as a way for publishers to block certain advertisers from placing ads in their apps. This is mostly used for 2 things: 1) blocking competitors and 2) blocking inappropriate ad content. This feature however, can also be used to force ad-networks into skipping ads that are being shown too much. If you focus on the top 5 ads shown in your app and only allow one ad-network to serve them you will force the other ones to bring new ads and diversify the user experience.

Getting more visibility to what ads are being shown

While a solution to this problem might look far fetched at the moment, it’s actually feasible. The ad-networks are under a lot of pressure to be more transparent at the moment and this is one area where if each network gives up some transparency it can receive a lot in return. After all, ad-networks also loose from ad blindness. It will be a better world for everyone, publishers, advertisers, mediation platforms, ad-networks and users. However, someone needs to take the first step. Until then, feel free to contact SOOMLA if you want access to this kind of information. A side benefit of publishers gaining access to this info is that it will accelerate the path to full transparency by the ad-networks.

Feel free to share:
App Monetization, Infographic

1BD277C0-AE65-4D4E-A593-28EED01006E2-5333-00000A48117CA031_tmp

This is the 3rd post in the series. The ad viewer of January is a user who made his app publisher very happy by generating the most amount of ad revenue for him when compared to other users. Our unique tech allows our customers to associate ad revenue to the user level and measure ad ltv. To check out previous months’ over achievers, follow this link for December and this one for November.

January Ad Viewer of the Month

This user alone generated 53.39 dollars for his app publisher in 20 activity days during January. What’s also interested is that he only recently started using the app – in mid December. The user contributed a bit over $20 in Dec. which makes his ARPU to date or his LTV to date $74. We are sure it will get even higher as he generates $2.66/day on average during January.

Attribue Ad Viewer of November
Country  United States
Device  iPad
Ad Types  Interstitial
Impresions  416
Active days 20
Revenue $53.39
eCPM $128.36
ARPDAU $2.66

IMG_4044

NOTE ABOUT SHARING – Feel free to share this infographic and embed it in your blog. If you do this, we will appreciate a link to http://soom.la.

Feel free to share:
App Monetization, Industry Forecasts

Latest report from App Annie supports the claim that the market is choosing View to Play as the business model of the future for the mobile ecosystem

Last November, while most of us were already preparing for the holidays, AppAnnie released a very interesting report that might have gone unnoticed by some of you. One of the Key Learnings is that Free-to-Play is giving way to View-to-Play. In other words, the fastest growing business model in the next 5 years in mobile apps will be in-app advertising and not in-app purchases.

In app advertising is growing at 24% CAGR and expected to surpass $110B by 2020 while Freemium is trailing behind

About App Annie and the report

The App Market data company needs no introduction from us and has become the source of data for most of the industry with regards to app store data. The company has over 600 employees in over 13, many of which are focused on researching data. From time to time, App Annie generates industry reports and forecasts and shares those through it’s blog and other content channels.
Company website – https://www.appannie.com/
Report Download Page – http://go.appannie.com/report-app-annie-app-monetization-2016-dg

What is View to Play?

If you haven’t heard the term View to Play, it’s probably because it’s new. When the app store just emerged, apps were sold and not given away for free. With the introduction of In-App Purchases, developer quickly started offering free apps to attract more users and find different ways to monetize them. This led to a new breed of game companies that specializes in conversion optimization, analytics, segmentation and performance marketing – the term Games as a Services was coined to reflect these new practices as well as Free to Play gaming. View to Play is similar in approach but instead of pushing users towards in-app purchases, the optimizations are focused around ad based monetization models – hence, “View to Play”. Users who want to advance in the game are often offered rewards and incentives for watching ads and a new breed of companies emerges with a toolset that includes special analytics capabilities around ad revenue measurement.

What is Driving the Change

In a recent article we covered how CPI is increasing and companies needs to adapt quickly. Well, some have already started and the App Annie report hints that more companies will be adopting the view to play model in the future.

These companies are realizing something that others have not. The CPI increase is highly correlated with the expected increase in ads LTV. They are both been driven by the same forces – the total mobile advertising spend is increasing twice as fast as the user growth. IAP revenues are actually increasing slower than the user growth and are becoming more and more concentrated in top grossing apps.

The cost per install is increasing over time as well as the average ad based revenue per user while In-App Purchase models are declining

This means that companies who transition quickly to view to play will be far better prepared for the future increase in CPIs. That is, as long as they can also adapt their measurement and optimization practices with a platform such as SOOMLA Traceback.

Feel free to share:
App Monetization, Infographic

Ad viewer of December is the user who generated the most amount of ad revenue for his app publisher

We are continuing the series of Ad Viewer of the month that we started last month. This type of analysis is one of the things that sets SOOMLA apart. We are using the Traceback technology to provide publishers with reports that get as granular as a single user. The Ad Viewer of December is a single user who made the most amount of revenue for the publisher of the apps he was using. Here is the link for last month’s report – Ad Viewer of November

December Ad Viewer of the Month

The amount of ad revenue generated by this user is mind blowing – $52.92 generated for the app publishers. He registered 19 active days in the month of December and made an average of $2.78 in each one of them. Unlike the Ad Viewer of November, this user also received a lot of in-game rewards for his revenue contribution. His favorite ad-types were Offer Wall and Rewarded Video that surely gave him incentives for his ad interactions.

Attribue Ad Viewer of November
Country  United States
Device iPhone
Ad Types  Offer Wall, Rewarded Video
Impresions  398
Active days 19
Revenue $52.92
eCPM $132.98
ARPDAU $2.78

Infographic featuring the ad viewer of December and different attributes about him. How much revenue was generated and at what eCPM

NOTE ABOUT SHARING – Feel free to share this infographic and embed it in your blog. If you do this, we will appreciate a link to http://soom.la.

Feel free to share:
App Monetization, Video

Blog header image - what's inside the advertising black box - video snaps from casual connect panel
Last Casual Connect in Tel-Aviv introduced many interesting lectures and panels. However, this is the one when ad-networks secrets got revealed. These are the top 9 moments of the panel presented in an easy video navigation tool.

 

Panel Participants

Lior Shiff – Co-founder and ex-CEO, Product Madness

Guy Tomer – Co-founder and CMO, TabTale

Niko Vouri – Co-founder and COO, Rocket Games

Yaniv Nizan – Co-founder and CEO, SOOMLA

Noam Neuman – VP Mobile Strategy at Matomy

Fernando Pernica – Mobile Monetization at Ad-Colony

Minute 5:29 – The Secret Guage

Lior asks Fernando whether there is a way for ad-networks to dynamically manipulate rev-share rates for publishers and create periods where they are more competative. Can you gues the answer?

Minute 12:06 – What Surprised Yaniv

Lior asks Yaniv what surprised him the most when lifting the hood of the black box. Not all app users are made equal apparently.

Minute 14:30 – When Ad Networks get Naughty

Guy tells the story about an ad-network that didn’t play by the rules and showed inappropriate ads to kids user audience.

Minute 32:11 – Brands – Friend or Foe

When a big change comes along you can either get defensive or find the opportunities that change creates. While the entrance of brands to mobile ads makes buying users harder it creates new monetization opportunities that translates back into the ability to place more competitive CPI bids.

Minute 33:09 – Is there an Unbiased Mediation?

Why is the ownership of mediaiton by ad-networks a problem? Bias and lack of transparency come into play here.

Minute 35:54 – Ad Networks’ Transparency

Guy explains that regardless of their various attempts to get more data from the ad-networks they still couldn’t get granular data and even aggregated data is sometimes tough.

Minute 37:07 – Lack of Transparency is a Double Edged Sword

Fernando explains how mediation is a black box for the ad-networks and how the lack of transparency goes both ways.

Minute 39:53 – Are There Ad Whales?

Lior is asking Yaniv and Guy whether or not Ad Whales exist. Guy explains that he can’t track it today but Yaniv is answering with precision: “We have seen $124 generated by a single user”.

Minute 45:43 – How Would You Leverage Ad LTV Data

Yaniv is asking Niko what would he do differently if he had the power to know. Niko explains how granular ad revenue data can impact their user acquisition decisions.

Feel free to share:

DOWNLOAD OUR RECENT REPORT

Join 7531 other smart people who get email updates for free!

We don't spam!

Unsubscribe any time

Categories

SOOMLA - An In-app Purchase Store and Virtual Goods Economy Solution for Mobile Game Developers of Free to Play Games